2
1.2 The Mobile Spatial coordinate Measuring System (MScMS)
This thesis introduces a new distributed measuring system called Mobile Spatial coordi-
nate Measuring System (MScMS), developed at the Industrial Metrology and Quality
Laboratory of DISPEA – Politecnico di Torino. MScMS has been designed to perform
simple and rapid indoor dimensional measurements of large-size objects (large scale me-
trology). It is made up of three basic parts:
• a “constellation” of wireless devices [MIT C.S.A.I.L., 2004];
• a mobile probe;
• a PC to store and elaborate data.
The wireless devices and the mobile probe use ultrasound (US) transceivers in order to
evaluate mutual distances implementing a Time of Arrival (ToA) technique. The mo-
bile probe localization in the working volume is obtained by a trilateration. The system
makes it possible to calculate the position – in terms of spatial coordinates – of the ob-
ject points “touched” by the probe. Acquired data are then available for different types
of elaboration to determine the geometric features of the measured objects (distances,
curves or surfaces).
This document focuses its attention on the description of the major challenges encoun-
tered while designing, implementing, and evaluating the MScMS system.
1.3 Organization of the dissertation
The remainder of this dissertation contains a detailed description of the MScMS working
principle. Then, the performances of the first MScMS prototype have been evaluated.
Three semi-automatic network localization procedures are described and compared. Fi-
nally all the diagnostic tools implemented are discussed. More specifically, the thesis is
structured as follows:
• Chapter 2 provides a detailed overview of large scale metrology systems. The attention
is focused on the already existing distributed systems highlighting their major advan-
tages and drawbacks.
• Chapter 3 presents the MScMS design features and working principle. In particular the
first MScMS prototype is described, presenting a preliminary experimental evaluation
of its metrological performance.
Introduction 3
• Chapter 4 and 5 deal with the problem of network localization. After a literature review
of existing localization algorithms for distributed wireless sensor networks, chapter5
proposes three different network localization procedures that have been tested on
MScMS.
• Chapter 6 concentrates the attention on the diagnostic tools implemented by the sys-
tem, describing them also by means of some examples.
Finally, chapter 7 summarizes the thesis contributions and mentions possible future direc-
tions for improving the MScMS performance.
2. Large scale metrology instruments, the new paradigm
of distributed systems
2.1 Introduction
The field of large scale metrology can be defined as “the metrology of objects in which
the linear dimensions range from tens to hundreds of meters” [Puttock, 1978]. Alterna-
tively, large-scale metrology could be said to apply to any dimensional measurement
where the metrology instrument has to be brought to the object rather than vice-versa
[Flack, 2005]. Over the recent years several innovative new large-scale metrology prod-
ucts such as laser radar, and indoor GPS (iGPS), as well as new methods for employing
metrology for example in complicated measurement networks have come to market.
These new technologies potentially offer end-users benefits over more mature technolo-
gies, for example, greater speed, higher point cloud density, more flexibility etc. While
these new technologies are attractive to potential end-users there are concerns that stan-
dards and independent performance verification techniques are not keeping up with such
technologies.
Beside classical centralized metrology systems, in which a stand-alone unit can provide
geometrical features of an object to be measured, latest advances in large scale metrology
are proposing distributed solutions in which a network of metrology stations share the
measurement task: a mobile sensor or target is localized relying on all the measurement
obtained by the metrology stations. The coming of such new metrology approach is the
result of a continuous effort toward scalable technologies able to cover flexible working
spaces. Classical centralized approaches to large scale metrology such as laser trackers,
theodolites or gantry CMMs can often result unhandy when dealing with complex work-
ing areas or dimensions of hundreds of meters. Distributed metrology systems, on the
contrary, can easily shape the working area by opportunely adding one or more metrology
stations. If flexibility and scalability are probably the major advantages of distributed me-
trology systems, their metrological performances are still hardly comparable to those of
centralized systems.
The aim of this chapter is to draw a picture of the major large scale metrology systems,
characterizing distributed systems by focusing the attention on their strengths and weak-
nesses. The remaining of this chapter is organized in four sections. Section 2.2 will
6
briefly describe all major large scale metrology instruments and their classification ac-
cording to their features. Distributed metrology systems will be described and character-
ized in detail in sections 2.3. Section 2.4 characterises distributed metrology systems ac-
cording to their major drawbacks and advantages.
2.2 Overview of large scale metrology systems
Maisano et al. proposed [Maisano et al., 2008] a classification of the major large scale
metrology instruments according to the following definitions:
• Centralized systems: a centralized system is essentially a stand-alone unit which can
work independently to provide the measurement of a spatial coordinate on the ob-
ject surface, e.g. a laser tracker. In some cases, a number of centralized systems can
be simultaneously used with the aim of improving the measurement accuracy.
• Distributed systems: a distributed system consists in a series of measuring stations
that work cooperatively to collect information for determining coordinates of a
point on the object’s geometry. In general, the individual stations associated with a
distributed system cannot measure coordinates separately.
• Contact systems: a contact measuring system is a metrology system able to provide
the coordinate of the object to be measured only by touching it with a probe. The
probe of the metrology system can be moved either manually or by mechanical
arms or can be attached to the object as a target to be followed by the system.
• Non contact systems: these systems are able to evaluate dimensional features of the
object to be measured without the need for a probe to touch the object. They are
mainly based on optical technologies.
Table 2.1 presents the major large scale metrology instruments classified according to the
proposed taxonomy.
Large scale metrology instruments, the new paradigm of distributed systems 7
CENTRALISED DISTRIBUTED
CONTACT
CMM,
Laser tracker,
Total station
iGPS,
Hi-Ball,
Contact systems with multilat-
eration technique
NON CONTACT
Theodolite, Laser radar,
Tacheometer,
Optical probe CMM,
Camera based triangulation
Photogrammetry,
Non-contact systems with multi-
lateration technique
Tab. 2.1 – Classification of major large volume measuring instruments [Maisano et al., 2008]
Large scale measuring systems presented in Table 2.1 can be further classified depending
on their working principle [Cuypers et al., 2008], as described below:
• Measuring systems that use two angles and one length. Most of the large-scale
measuring systems rely on the determination of one length and two angles. In these
systems the initial coordinates of a point are evaluated in a spherical coordinate sys-
tem (ρ, ϕ, θ). For this reason these systems are also called Spherical Coordinate
Measurement Systems. For each system, the angles are measured by means of an-
gular encoders, whilst the range measurement can be performed using either an in-
terferometer like laser trackers or an ADM (Absolute Distance Measure) like laser
radars and total stations, or a combination of both the two technologies like ADM
enabled laser trackers. The spherical coordinates are easily transformed in Cartesian
coordinates by a central computing unit that is able to derive the object features
from the points’ information.
Fig. 2.1 – Working principle of laser trackers, laser radars and total stations [Cuypers et al., 2008].
The position of a point is defined by a range distance and two angular measurements
8
• Measuring systems using multiple angles (triangulation). Instead of using two an-
gles and a distance measurement, it is possible to evaluate the position of a point in
a three dimensional space using just angular information from two or more refer-
ence points. This working principle relies on very well known triangulation algo-
rithms. Triangulation uses the known locations of two or more reference points, and
the relative angles between the point to be localized and each reference point. In
this case the unknown position of the point can be found by solving a linear system
[Dogancay et al., 2005].
The camera based triangulation system applies this principle. Three or more CCD
(Charge Coupled Device) cameras are placed in known positions, for example fixed
on a support. Each of them looks at a target determining the plane containing it. The
position of the target is then univocally determined as the intersection of all the
planes (see Fig. 2.2).
Fig. 2.2 – The working principle of camera based triangulation. The position of the target is de-
termined by planes intersection.
• The working principle of the iGPS is also based on multiple angle measurements.
Knowing the horizontal and vertical angles from two or more transmitters, the sys-
tem univocally determines the position of a probe. In order to obtain accurate angle
measurements the iGPS uses rotating laser beams [ARC Second, 2009]. Next sec-
tion will describe this system in more detail.
Photogrammetry is another large-scale measurement technique based on an-
gle measurements. The principle is similar to that of camera based triangula-
tion, but, in this case, camera positions are not precalibrated, but they can be
calculated afterwards together with the target position [Mikhail et al., 2001].
Large scale metrology instruments, the new paradigm of distributed systems 9
• Measuring system using multiple lengths (trilateration). Trilateration uses the
known locations of three or more reference points, and the measured distance be-
tween the point to be localized and each reference point (see Fig. 2.3). The un-
known coordinates can be found by solving a non linear optimization problem
[Franceschini et al., 2007]. This procedure is adopted for instance for CMM calibra-
tion, but it is quite expensive due to its requirement to at least three laser interfer-
ometers. Because of its cost, in the domain of large scale metrology this approach is
mainly used for research activity.
V
1
V
2
V
3
d
1
d
2
d
3
Fig. 2.3 – The working principle of trilateration in two dimensions. Knowing the distances from
three different reference points (V
1,
V
2
, V
3
) it is possible to localize the position of an unknown
point on the intersection of three circumferences.
2.3 Distributed metrology systems
As introduced in the previous sections, distributed systems consists of a series of measur-
ing stations that work cooperatively with the aim of determining object geometrical fea-
tures. In general, the individual stations associated with a distributed system cannot
measure coordinates. Sometimes also centralized metrology system can be used in a dis-
tributed way when for example multiple metrology facilities are available in the same fac-
tory.
Among all large scale metrology instruments classified above, the distributed instruments
are the most recent ones and for this reason they are the ones which show the greatest po-
10
tential and still being subject of research [Flack et al., 2005]. Although photogrammetry
is a mature and well known distributed technique, its real potential is still to be discov-
ered due to the recent advances in computer performance within the last decade [Mikhail
et al., 2001].
All the distributed systems have a common similar structure (see Fig. 2.4):
• A network of multiple components distributed around the working area.
• A wireless mobile probe equipped with sensors able to detect the signals received
from the distributed components or a set of target sensors to be localized. The mo-
bile probe can be moved either manually or automatically by a robot.
• A computing unit able to process the data streaming sent by the mobile probe or the
constellation components.
Distributed
network
Mobile
probe
Computing
unit
Measured
object
Fig. 2.4 – Classical layout of a distributed metrology system.
All distributed metrology systems rely on two of the described working principles: de-
pending on the measurement capability of the measurement stations, distributed systems
can measure using multiple angles or lengths. In the following, the four more relevant
distributed metrology solutions are described.
Hiball
The HiBall Tracker is a new approach to wide-area tracking and measuring, delivering
accuracy with low latency, high update rate, and scalability to cover a large region. It is
composed of two key integrated components; the HiBall Optical Sensor mounted on a
mobile probe and the HiBall Ceiling Beacon Arrays (the constellation components). The
Large scale metrology instruments, the new paradigm of distributed systems 11
HiBall Optical Sensor is composed of 6 lenses and photodiodes arranged so that each
photodiode can ‘view’ infrared LEDs (Light Emitting Diodes) of the Beacon Arrays
mounted on the ceiling, through several of the 6 lenses (see Fig. 2.5) [Welch et al., 2001].
Mobile probe
Constellation components
Fig. 2.5 – The HiBall system. The constellation components are embedded in a series of ceiling
mounted strips while the probe is equipped with photodiodes sensors.
The mobile probe is able to estimate the angles from the HiBall Beacon Arrays seen by
the photodiodes. The position of the probe is found by triangulation knowing the loca-
tions of the Beacon Arrays. Although the localization of the HiBall Beacon Arrays is
needed by the system, no special adjustments are required of the ceiling structure — the
system’s precision is unaffected by typical variations in ceiling height or in strip place-
ment. The HiBall system's auto-calibration feature then rapidly determines the location
of the individual ceiling strips creating a “ceiling map” that can be saved and/or continu-
ously updated while tracking [Welch and Bishop, 1997].
This system works at a frequency of about 2000 Hz, suitable also for accurate tracking
even with rapid movements. The location resolution is higher than 0.2 mm [Welch et al.,
2001] with an angular accuracy of more than 0.01 degrees.
The iGPS
Among all distributed contact metrology systems, the indoor GPS (iGPS) is probably the
most advanced one, whilst the others are still object to development, the iGPS is already
commercialized and used in some environmental contexts. The system components of
Ricevi informazioni sui nostri servizi, sulle offerte e non perdere news e consigli su università e lavoro.
